News for hackers. Stuff that matters.
Re:ARMv7 vs. x86-64: Pathfinding benchmark of C++, D, Go, Ni... by logicchains on Saturday December 20, 2014 @06:59AM (#8776430)
>To the best of my knowledge Typed Racket does support gradual typing

If you take a Racket program and change the language to Typed Racket, you'll get flooded with compiler errors. Or at least that's been my experience.

>Racket is not Scheme anymore

I'll update the post to note this, thanks.

Re:ARMv7 vs. x86-64: Pathfinding benchmark of C++, D, Go, Ni... by klibertp on Saturday December 20, 2014 @10:34AM (#8776843)
> If you take a Racket program and change the language to Typed Racket, you'll get flooded with compiler errors. Or at least that's been my experience.

That's because when using `#lang typed/racket` you literally change the language you're writing and that includes the syntax of some very fundamental forms (like define or struct).

You can mix typed and untyped code freely as long as they are separated by module boundaries. Something like this:

    #lang racket/base

    (module with-types typed/racket
      (provide a)
      (: a (-> String Integer))
      (define (a x) 0))

    (module without-types racket
      (require (submod ".." with-types))
      (provide a))
But it's definitely not as convenient as Erlang's Dialyzer, TypeScript, Dylan or similar success/gradual/occurrence/soft-typing solutions for sure.
Re:ARMv7 vs. x86-64: Pathfinding benchmark of C++, D, Go, Ni... by kuschku on Saturday December 20, 2014 @10:24AM (#8776818)
You can use typed/racket with optional types, it’s kinda easy.
Re:ARMv7 vs. x86-64: Pathfinding benchmark of C++, D, Go, Ni... by klibertp on Saturday December 20, 2014 @10:38AM (#8776854)
How? The docs say (http://docs.racket-lang.org/ts-guide/more.html#(part._when-a...) that you have to provide type annotations in certain cases.